THE ROLE OF THE DIFFERENT CITIES AND TOWNS
11 THE IMPACT OF PUBLIC POLICY ON MIGRATION AND EMPLOYMENT
In Chapter 2 the colonial policy of concentrating centers of economic growth in a limited number of geographic locations was described briefly. At the conclusion of the chapter the observation was made that the strong commitment to achieving economic growth in the postindependence period has kept the colonial economic structure basically intact. The migration survey results, as reported and analyzed in the subsequent chapters, have shown rural-to-urban migration to be a rational response by some members of the rural labor force t o this geographic concentration of economic opportunities.
In this concluding chapter attention is directed t o the aspects of the migration pro- cess that are amenable t o public policy. First, we develop further the precise nature of the aspects of the existing economic structure that shape and determine the migration process. Second, a variety of policy options that can directly affect the migration process are presented and the likely effectiveness of each is assessed. Finally, some attention is devoted to additional, more radical policies that could change the aspects of the existing economic structure that are the underlying causes of the current internal migration.
THE STRUCTURE OF THE URBAN ECONOMY
As summarized in Chapter 2, the Kenyan economy has displayed commendable advances since independence. First, the annual rate of growth of gross domestic product through 1968 was just over 6 percent. This rate of growth has been maintained into the 1970s (Development Plan: 1974-1978 (1974), Vol. I, p. 148). In addition, entrants into secondary schools have more than doubled and the number of secondary school-leavers has more than tripled (Kinyanjui 1973, p. 79).
Less commendable is the unequal distribution of benefits generated by these advances in the economy. The scanty evidence available indicates that most of the benefits accrue to a small minority: approximately 13 percent of the households obtain 72 percent of the annual income flow generated in the economy (Rempel 1975, Table 1). This minority includes owner-operators of large-scale businesses and farms, slulled and some semiskilled employees, and the owners of some of the smaller businesses and cash crop farms.
Their position in the economy provides this minority with several important privi- leges: they control the important means of production; they claim virtually all of the rent and operating surplus generated in the economy; they dominate the demand for output and hence largely determine what goods will be produced in the modem sector.
Given the taste preferences of this wealthy minority, the type of goods and services demanded frequently require technology and capital beyond the immediate capability of existing or potential Kenyan businessmen.* To overcome import constraints, foreign firms
*The effect of this concentration of demand for particular goods and services is reinforced by the emphasis placed on tourism in Kenya's development strategy.
have been invited to supply the required technology and capital. Protection against imports was provided as an inducement where necessary. An additional inducement provided such firms, as a means of obtaining and retaining their services, was the provision o f a particular type of infrastructure in certain urban localities. This infrastructure causes other firms and government to concentrate their activities in these same localities. The combination of these forces has served t o agglomerate most modern sector employment in a selected number of urban centers.
For such foreign firms with a limited capacity to manage a diverse local labor force, the type of technology that has minimal labor requirements will be preferred. For this type of technology, high labor turnover increases significantly the risk of being able to maintain a desired level of profits over time. (A foreign-owned firm, relying on a transfer- pricing mechanism to repatriate profits t o a parent company, seeks t o minimize the vari- ance in profits over time as well as t o maximize the level of profits. The transfer-pricing mechanism does not lend itself t o repatriating widely fluctuating levels of profit.) To obtain the desired labor force stability a firm may well find it advantageous to invest some current profits in higher wages t o stabilize a desired level of profits over time.
Where the initial wage increase generates a "reserve army" of visibly unemployed at the factory gates, the firm provides the current employees with a powerful incentive t o remain at their jobs. (According to Mr. D. Richmond, Chief Executive Officer of the Federation of Kenya Employers, labor turnover in most of the firms represented by the federation had dropped below 5 percent a year by 1967168.) Once labor force stability has been accomplished, the firm will continue t o pay above the wage necessary t o obtain a desired supply of labor because of the increased labor efficiency made possible by low labor turnover. For a detailed discussion of this concept see Rempel and House (1978, pp. 78-81).
To initiate this process a firm must have sufficient financial resources to purchase the advanced technology and t o pay the initial higher wages. Provided the labor market is competitive, any one firm will not know the exact amount of the initial premium it has t o pay t o purchase labor force stability. As a result, a number of firms, and possibly gov- ernment, may well bid up wages beyond the level required for their initial purposes. For the firms that can pass such cost increases on in the form of higher prices and for govern- ment which can pass the higher wage costs on in the form of higher taxes, such initial trial- anderror mistakes need not present a serious problem.
For the purpose of subsequent analysis we will refer t o employment in such firms and in government as "sheltered" employment. The employed are sheltered in the sense that the firm pays a wage above what would be necessary t o induce an adequate labor sup- ply. The firm pursues this wage policy to obtain a stable labor force so its employees are protected against the unemployed lined up outside the firm's gates. This protection does not reflect control by employees over their work conditions; employers are not beyond using these potential replacements as a means of controlling their present employees.
The remaining modern sector firms, with limited financial resources and facing effective competition in the product market, will be forced out of the game of purchasing a stable, more efficient, selected labor force. As a result, employment in these firms will be labeled "competitive" employment. In practice the two employment lotteries cannot be identified as separate entities but it was useful to separate them conceptually. The bal- ance of the urban labor force will either need to seek some form of employment in the informal sector or will remain unemployed.
In Figure 11 .I an attcmpt is made to summarize this information as a set of employ- ment prospects that confronts urban in-migrants. The urban "pull" is seen to be the "shel- tered" employment and, possibly for the unskilled, the "competitive" employment. The hiring process, based heavily on education as a means of selection and influenced signifi- cantly through kin already employed, serves to channel where the in-migrants go. Fvi those who go the employment lottery routes, realization of employment (level 4) can take some time so kin also provide support during this time of job search. The dual roles of kin already resident in town provide the migration process withits own internal momen- tum: the more people there are in town the greater the potential to assist new entrants.
The informal sector is portrayed as primarily a residual employer rather than the place where in-migrants queue for the available modern sector jobs. Some of the success- ful entrepreneurs in the informal sector are voluntary drop-outs from "competitive"
employment. At level 6 in the flow diagram the urban income, Y,, is t o be seen as some- thing that can change dramatically as employment status changes. Hence, a layoff or an inability t o work because of a disability may be the immediate cause of the urban out- migration shown in the flow diagram. For other participants in the urban labor market, Y, is to be viewed also as a level of income that is compared by the migrant with the level of income perceived t o be available in his home area or in some other potential migration destination.
POLICY DEALING WITH SYMPTOMS OF MIGRATION
Established urban residents are notably concerned about the extent of urban in- migration because excessive numbers coming in can be a threat t o the way of life of the urban residents. Therefore, policy designed t o control in-migration is not merely being considered but in some cases is being put into effect in Kenya. We examine the basis for these policies in this section. It is recognized that such policies may have some immediate effect on the extent of in-migration but in most cases they deal with symptomsrather than underlying causes. Policies attacking the causes are addressed in the concluding section.
The Availability and Content of Education
The survey and census data analyzed in this study indicate a definite positive correla- tion between the level of formal schooling completed and the propensity t o migrate t o an urban location. In Chapter 5 we reported that the extent of the gap in the proportion that had completed a given level of schooling between the migrant sample and the total rural male population varied directly with the level of schooling completed. Further, we reported that the completion of a particular education stream served t o affect the timing of the rural-to-urban move. The regression results reported in Table 4.1 indicate a signifi- cant, positive relationship between the proportion of school-aged children in a district with formal education and the odds of a rural-to-urban move.
This evidence suggests that reductions in expenditures on formal schooling might reduce the extent of rural-to-urban migration, at least temporarily. Whether such a policy could be implemented is another matter. Given the role of education as a prerequisite for
participating in the "sheltered" employment lottery, the stimulus t o obtain formal school- ing is very strong for many young people and parents. According t o the government of Kenya, the rapid increases in secondary school enrollment "
.
. . have as their basis the conviction, on the part of almost all Kenyans, that education is very nearly the only ladder t o high status, t o good jobs and associated high incomes" (Sessional Paper on Employ- ment 1973, p. 47).Therefore, the starting point for policy change with reference t o education will have to be the "sheltered" employment: both in the level of wages paid and in the use of education as the primary selection device. Any reduction in the attractiveness of "shel- tered" employment would reduce the role that primary education now plays as the means t o advancement.
Such a change would have several beneficial effects o n the economy. First, the pro- jected increase during the period 1975 t o 1985 of school-leavers who will not be absorbed by wage employment is from approximately 150,000 t o 500,000 (Federation of Kenya Employers 1976, Table 13A). Any decline in the enrollment of people who d o not finish school (people who use government education funds b u t d o not contribute t o develop- ment) will better enable the government t o achieve its stated objective of reducing educa- tion costs from 1 5 percent t o 11 percent of its annual expenditures.
Second, the incentive of people t o repeat standard VII in the hope of obtaining sec- ondary school admission would be reduced. The potential reduction in wasted resources here is substantial. A survey in one district reported that as many as 7 8 percent of the students who failed to gain admission into a secondary school repeated standard VII (Somerset 1973, p. 27). The median time for completing the 7-year program was 8.2 years.
Third, changes in the need for secondary education for an opportunity for rapid personal advancement would open the possibility for redirecting the tremendous commun- ity energies and substantial local resources that have gone into building Harambee (self- help) secondary schools. These schools have become lowquality carbon copies of the government-supported secondary schools. The result is an educational system which is
"
.
..
not only producing the least needed type of manpower b u t also of low academicquality" (Kinyanjui 1973, p. 96). The community development potential of an altema- tive use of these resources and energies could be substantial.
Finally, such a proposed change in the incentive system could open the door for effective curriculum reform. The content of the existing curriculum is characterized by what doctors, engineers, teachers, economists, and administrators will need t o know, n o t by what is useful for rural development (Somerset 1973, p. 37). The premium placed o n rote learning t o obtain high scores o n exams given a t the end of each educational stream is generating graduates who lack both the skills and the selfconfidence t o venture o u t on their own in some form of self-employment (Kinyanjui 1 9 7 3 , p. 9 5 ; Federation of Kenya Employers 1976, p. 48).
The Federation o f Kenya Employers states (1976, p. 93): "The content of educa- tion even at present, bears little relevance t o the economic needs of the country, especially the rural areas." At another point (p. 39), they argue: " Employers, students and parents have n o significant control over curricula."
I t is our contention here that the federation, in the form of those of its members providing "sheltered" employment, does indeed determine the content of the current cur- riculum. Further, the educational system is serving these firms well. I t provides a suitable
employnlent screening device and it generates secondary school graduates who have devel- oped desired work habits and acquired some desired academic skills. The potential for using public policy to change the availability and content of education to decrease rural- to-urban migration is constrained severely by the actions of those firms that are operating the "sheltered" employment system.
E x p a ~ s i o n of Urban versus Rural Employment Opportunities
According to the migrants surveyed the spatial distribution of the employment opportunities is at the crux of the migration problem (see Tables 4.5 and 4.6). As a result, one would expect, cetems panbus, that any job creation projects in the towns will draw in more migrants since the odds of winning in the urban employment lottery will rise.
Given that the rate of urban in-migration is correlated positively with the rate of urban job creation and that the odds of a rural-urban move vary inversely with our mea- sure of urban employment prospects, we expect that the labor supply elasticity with respect to the urban job creation is larger than one.* As we indicate in Chapter 4 , our measure of urban employment prospects probably is capturing the increased unemploy- ment caused by the surge of in-migrants who are responding to the new employment opportunities. This was borne out by the labor supply response to the employment crea- tion inherent in the 1964 and the 1971 Tripartite agreements.
With reference to the other half of the policy question, whether expanding rural employment will serve to stem the extent of rural-to-urban flows, there is some debate.
According to the migrants the answer would be in the affirmative. Seventy-nine percent said that they would prefer t o live in their rural home areas provided comparable employ- ment to that in town was available there (based on responses to Question 46 as analyzed in Chapter 9). Further support for the position taken by the respondents is evident in the fact that the bulk of the new economic opportunities have been in the rural areas. These new opportunities are serving to absorb the majority of the new entrants into Kenya's labor force (Federation of Kenya Employers 1976, p. 118).
Evidence to the contrary hinges on the adequacy of these rural options relative to the employment opportunities available in the towns. The Federation of Kenya Employers estimates that approximately 3 0 percent of the rural households d o not have access to land (1976, p . 188). This is similar to our finding that 36 percent of the migrants were from landless households. In addition, the Federation of Kenya Employers argues that approximately one-fourth of the small-holder farmers do not have access to sufficient land t o be able t o produce a subsistence income from farming (1976, p. 8).
Some of these rural poor can be expected t o look t o the towns for a means of sur- vival. Our flow diagram (Figure 11 . l ) would locate them primarily in the informal sector even if income available there is very low. Evidence that the landless are concentrated in informal activities in the urban slums is provided by Ross (1973, p . 143) in his study of Mathare Valley in Nairobi.
In addition to the rural poor there are those households which may well have an income above subsistence but it is too low to enable them to fulfill their aspirations
(9).
*In an unpublished paper Scott (1972) provides a set of assumptions and a range of parameter values for Kenya which would generate a labor supply elasticity less than one.
For these llouseholds i t is n o t merely a matter that rural employment opportunities are available; if a rural-to-urban move is t o be prevented the level of income generated by the employment must be comparable t o what household members perceive t o be available t o them in town.
Some rural employers find i t difficult t o recruit an adequate labor supply. Thus fact.
together with the fact that open unemployment is evident in the towns, indicates that n o t all rural employment opportunities provide a level of remuneration comparable t o what people hope t o obtain elsewhere. The Federation of Kenya Employers attributes these rural labor shortages t o the perverse effect o f education which turns people against farm labor (1976, p. 39). In contrast, both Gwyer (1972, p . 4 ) and Wasow (1973, p. 25) attri- bute these labor shortages t o the low wages and poor working conditions inherent in these jobs. F o r the better educated at least, m o d e m sector employment in town is probable enough that they can exercise some choice with reference t o where they will work.
The creation of new employment opportunities in rural areas is a policy option under direct control of the government. First, it can exercise some control over the spatial location of its activities. In addition, the Ministry of Agriculture can be instrumental in the promotion of the type of farming techniques which increase labor productivity but minimize the substitution of capital for labor. The Johnston and Kilby (1975) unimodel approach t o agricultural development, with backward linkages t o the type of agricultural inputs that can be produced within Kenya without primary reliance o n outside sources of technology, capital, and entrepreneurship, would be an appropriate model t o pursue.
Where such rural development efforts are implemented it must be recognized that the aspiration levels of the households involved may also be raised by these development projects. If so, these projects may induce rural-to-urban migration if the increased rural income generated by the projects is less than both the new aspiration levels of the house- holds and the income they perceive t o be available to their members in one of the urban centers.
Changes in Urban versus
Rural
Wage LevelsThe initial extension of the Todaro migration model by Harris and Todaro (1970) advocated an urban wage subsidy as a means for absorbing the urban in-migration. Although they did n o t consider all the implications of such a subsidy (see Stightz 1974), and their position was premised on the unrealistic assumption that modern sector labor-capital ratios are quite responsive t o changes in relative factor prices, this article did serve t o focus attention on wage policy as the prime mechanism for affecting the migration process.
The initial extension of the Todaro migration model by Harris and Todaro (1970) advocated an urban wage subsidy as a means for absorbing the urban in-migration. Although they did n o t consider all the implications of such a subsidy (see Stightz 1974), and their position was premised on the unrealistic assumption that modern sector labor-capital ratios are quite responsive t o changes in relative factor prices, this article did serve t o focus attention on wage policy as the prime mechanism for affecting the migration process.